The Metal-Rich Nature of Stars with Planets?

A&A 373, 1019–1031 (2001) Astronomy DOI: 10.1051/0004-6361:20010648 & c ESO 2001 Astrophysics

The metal-rich nature of stars with planets?

N. C. Santos1, G. Israelian2, and M. Mayor1

1 Observatoire de Gen`eve, 51 ch. des Maillettes, 1290 Sauverny, Switzerland 2 Instituto de Astroﬁsica de Canarias, 38205 La Laguna, Tenerife, Spain

Received 13 March 2001 / Accepted 3 May 2001

Abstract. With the goal of confirming the metallicity “excess” present in stars with planetary-mass companions, we present in this paper a high-precision spectroscopic study of a sample of dwarfs included in the CORALIE extrasolar planet survey. The targets were chosen according to the basic criteria that 1) they formed part of a limited volume and 2) they did not present the signature of a planetary host companion. A few stars with planets were also observed and analysed; namely, HD 6434, HD 13445 (Gl 86), HD 16141, HD 17051 (ι Hor), HD 19994, HD 22049 ( Eri), HD 28185, HD 38529, HD 52265, HD 190228, HD 210277 and HD 217107. For some of these objects there had been no previous spectroscopic studies. The spectroscopic analysis was done using the same technique as in previous work on the metallicity of stars with planets, thereby permitting a direct comparison of the results. The work described in this paper thus represents the first uniform and unbiased comparison between stars with and without planetary-mass companions in a volume- limited sample. The results show that 1) stars with planets are significantly metal-rich, and 2) that the source of the metallicity is most probably “primordial”. The results presented here may impose serious constraints on planetary system formation and evolution models.

Key words. stars: abundances – stars: chemically peculiar – planetary systems

1. Introduction Opposing to this view, it had been proposed that the observed metallicity “excess” may be related to the “pol- One of the most exciting and promising results that be- lution” of the convective envelope of the star by the in- came evident after the discovery of the first extrasolar fall of planets and/or planetesimals (e.g., Gonzalez 1998; planets is that their host stars seem to be very metal- Laughlin & Adams 1997; Laughlin 2000; Gonzalez et al. rich when compared with dwarf stars in the solar neigh- 2001). This pollution can be the result of the “complete” bourhood (Gonzalez 1998; Santos et al. 2000a – hereafter inward migration of a planet on to the star, the transfer Paper I; Gonzalez et al. 2001). This result, representing of material from the disc to the star as a result of the the only link between the presence of planets and a stellar migration process (Goldreich & Tremaine 1980; Lin et al. photospheric feature, has been given two main explana- 1996), or to the break-up and infall of a planet(s) on to tions. The first, is based on the classical view that giant the surface of the star due to gravitational interactions planets are formed by runaway accretion of gas on to a with other companions (Rasio & Ford 1996). The former “planetesimal” having up to 10 Earth masses. In such a point might be particularly important for the short-period case, we can expect that the higher the proportion is of systems (Queloz et al. 2000). dust to gas in the primordial cloud (i.e. metals), and con- The idea that planets and/or planetary material might sequently in the resulting proto-planetary disc, the more be engulfed by a star is to some extent supported by the rapidly and easily may planetesimals, and subsequently recent detection of 6Li in the planet host star HD 82943 the now observed giant planets be built. (Israelian et al. 2001) – interpreted as a signal of the accretion of a planet during the history of the star – and prob- Send offprint requests to:N.C.Santos, ably by the detection of a significant difference in iron e-mail: [email protected] abundances in the very similar pair of dwarfs 16 Cyg A ? Based on observations collected at the La Silla Observatory, and 16 Cyg B, which also have very different Li contents ESO (Chile), with the CORALIE spectrograph at the 1.2-m (Laws & Gonzalez 2001; Gonzalez 1998). Euler Swiss telescope, with the FEROS spectrograph at the 1.52-m ESO telescope (Observing run 66.C-0116 B), and using Besides the [Fe/H] differences, there is currently some the UES spectrograph at the 4-m William Hershel Telescope debate about possible anomalies concerning other ele- (WHT), at La Palma (Canary Islands). ments (Paper I; Gonzalez et al. 2001; Smith et al. 2001).

Article published by EDP Sciences and available at http://www.aanda.org or http://dx.doi.org/10.1051/0004-6361:20010648 1020 N. C. Santos et al.: The metal-rich nature of stars with planets

Table 1. Volume-limited sample of stars without detected giant planets.

Star Teﬀ log gξt [Fe/H] N(Fe i) N(Fe ii) σ(Fe i) σ(Fe ii)Mass −1 (K) (cgs) (km s )[M ]

HD 1581 5940 4.44 1.13 −0.15 31 7 0.04 0.05 0.99 0.02 HD 4391 5955 4.85 1.22 0.01 36 5 0.05 0.09 1.22 0.04 HD 5133 5015 4.82 0.92 −0.08 36 6 0.05 0.07 0.81 0.04 HD 7570 6135 4.42 1.46 0.17 35 7 0.04 0.06 1.19 0.02 HD 10360 5045 4.77 0.89 −0.19 36 5 0.04 0.04 0.70 0.03 HD 10647 6130 4.45 1.31 −0.03 34 7 0.03 0.05 1.14 0.03 HD 10700 5370 4.70 1.01 −0.50 38 6 0.04 0.04 0.68 0.02 HD 14412 5410 4.70 1.01 −0.44 35 6 0.04 0.01 0.78 0.05 HD 17925 5220 4.60 1.44 0.08 35 6 0.07 0.04 0.92 0.06 HD 20010 6240 4.27 2.23 −0.20 33 6 0.05 0.08 1.33 0.01 HD 20766 5770 4.68 1.24 −0.20 35 7 0.04 0.04 0.97 0.04 HD 20794 5465 4.62 1.04 −0.36 39 7 0.05 0.05 0.74 0.02 HD 20807 5865 4.59 1.28 −0.22 37 7 0.04 0.04 0.95 0.02 HD 23249 5135 4.00 1.12 0.17 36 7 0.06 0.08 0.84 0.01 HD 23356 5035 4.73 0.96 −0.05 36 6 0.06 0.07 0.83 0.03 HD 23484 5230 4.62 1.13 0.10 37 6 0.05 0.07 0.92 0.06 HD 26965A 5185 4.73 0.75 −0.26 37 5 0.05 0.03 0.71 0.02 HD 30495 5880 4.67 1.29 0.03 37 7 0.04 0.03 1.11 0.04 HD 36435 5510 4.78 1.15 0.03 37 6 0.06 0.03 1.04 0.05 HD 38858 5750 4.56 1.22 −0.22 36 7 0.04 0.02 0.91 0.02 HD 39091 5995 4.48 1.30 0.09 37 7 0.04 0.04 1.10 0.02 HD 40307 4925 4.57 0.79 −0.25 37 4 0.06 0.10 0.76 0.07 HD 43162 5630 4.57 1.36 −0.02 35 7 0.05 0.04 0.99 0.04 HD 43834 5620 4.56 1.10 0.12 38 7 0.05 0.05 0.96 0.02 HD 50281A 4790 4.75 0.85 0.07 31 4 0.05 0.12 0.79 0.01 HD 53705 5810 4.40 1.18 −0.19 36 7 0.03 0.03 0.92 0.01 HD 53706 5315 4.50 0.90 −0.22 36 7 0.05 0.08 0.88 0.05 HD 65907A 5940 4.56 1.19 −0.29 39 7 0.04 0.04 0.94 0.01 HD 69830 5455 4.56 0.98 0.00 38 7 0.05 0.03 0.89 0.06 HD 72673 5290 4.68 0.81 −0.33 38 6 0.04 0.04 0.78 0.06 HD 74576 5080 4.86 1.20 0.04 36 5 0.06 0.05 0.91 0.04 HD 76151 5825 4.62 1.08 0.15 38 7 0.03 0.04 1.09 0.04 HD 84117 6140 4.35 1.38 −0.04 34 7 0.04 0.06 1.14 0.01 HD 189567 5750 4.57 1.21 −0.23 37 7 0.04 0.05 0.89 0.01 HD 191408A 5025 4.62 0.74 −0.51 37 4 0.06 0.09 0.61 0.04 HD 192310 5125 4.63 0.88 0.05 36 6 0.06 0.08 0.78 0.04 HD 196761 5460 4.62 1.00 −0.27 38 7 0.05 0.05 0.81 0.03 HD 207129 5910 4.53 1.21 −0.01 36 7 0.04 0.03 1.03 0.02 HD 209100 4700 4.68 0.60 0.01 34 3 0.07 0.06 0.64 0.09 HD 211415 5925 4.65 1.27 −0.16 35 7 0.03 0.04 0.99 0.02 +0.01 HD 216803 4647 4.88 0.90 0.07 28 3 0.07 0.08 0.77−0.18 HD 222237 4770 4.79 0.35 −0.22 37 3 0.08 0.08 – HD 222335 5310 4.64 0.97 −0.10 33 5 0.05 0.04 0.86 0.05

But the relatively low number of exoplanets known, and hosting planets with those of stars “without” planets, au- possible systematics with respect to the samples do not thors had access only to published metallicity studies of permit ﬁrm conclusions to be reached on the subject. volume-limited samples of dwarfs in the solar neighbour- The observation that stars with planets1 are particu- hood (mainly that of Favata et al. 1997, the alternative larly metal-rich has so far been overshadowed by the re- being to construct a sample using less precise metallici- striction that in order to compare the metallicities of stars ties using photometric indices). This is inconvenient for a number of reasons. First and most obviously, one cannot 1 Here we are referring to the known extrasolar planetary be sure if the sample used for comparison is really free host stars, having Jupiter-like planets in relatively short period from giant planets. Furthermore, the metallicities for the orbits when compared to the giant planets in our own Solar Favata et al. sample were determined using a much shorter System. N. C. Santos et al.: The metal-rich nature of stars with planets 1021

Table 2. Determined atmospheric parameters and [Fe/H] for a set of stars with planets and brown dwarf companions (bd).

Star Teﬀ log gξt [Fe/H] Instrument Mass Planet −1 (k) (cgs) (km s )[M ]discoverypaper

HD 1237 5555 50 4.65 0.15 1.50 0.08 0.11 0.08 CORALIE 1.01 0.04 Naef et al. (2001a) HD 6434 5790 40 4.56 0.20 1.40 0.10 −0.55 0.07 FEROS 0.79 0.01 Queloz et al. (2001) HD 13445 5190 40 4.71 0.10 0.78 0.10 −0.20 0.06 CORALIE HD 13445 5220 40 4.70 0.10 0.85 0.07 −0.19 0.06 FEROS HD 13445 (avg) 5205 4.70 0.82 −0.20 – 0.75 0.04 Queloz et al. (2000) HD 16141 5805 40 4.28 0.10 1.37 0.09 0.15 0.05 FEROS 1.06 0.01 Marcy et al. (2000) HD 17051 6225 50 4.65 0.15 1.20 0.08 0.25 0.06 FEROS 1.26 0.02 K¨urster et al. (2000) HD 19994 6165 40 4.13 0.20 1.49 0.10 0.23 0.06 CORALIE HD 19994 6250 40 4.27 0.10 1.56 0.08 0.30 0.06 FEROS HD 19994 (avg) 6210 4.20 1.52 0.26 – 1.34 0.01 Queloz et al. (2001) HD 22049 5135 40 4.70 0.10 1.14 0.07 −0.07 0.06 CORALIE 0.85 0.04 Hatzes et al. (2000) HD 28185 5705 40 4.59 0.10 1.09 0.06 0.24 0.05 CORALIE 0.98 0.07 Mayor et al. (2001b) HD 38529 5675 40 4.01 0.15 1.39 0.09 0.39 0.06 FEROS 1.52 0.05 Fisher et al. (2000) HD 52265 6075 40 4.21 0.10 1.31 0.07 0.22 0.07 CORALIE HD 52265 6120 50 4.37 0.20 1.31 0.06 0.25 0.06 FEROS HD 52265 (avg) 6100 4.29 1.31 0.24 – 1.18 0.01 Naef et al. (2001a) Butler et al. (2000) HD 75289 6135 40 4.43 0.20 1.50 0.07 0.27 0.06 CORALIE 1.20 0.02 Udry et al. (2000) HD 82943 6025 40 4.54 0.10 1.10 0.07 0.33 0.06 CORALIE 1.15 0.05 Naef et al. (2001a) HD 83443 5500 60 4.50 0.20 1.12 0.09 0.39 0.09 CORALIE 0.90 0.05 Mayor et al. (2001a) HD 108147 6265 40 4.59 0.15 1.40 0.08 0.20 0.06 CORALIE 1.27 0.02 Pepe et al. (in prep.) HD 121504 6090 40 4.73 0.10 1.35 0.08 0.17 0.06 CORALIE 1.19 0.03 Queloz et al. (2001) HD 162020bd 4830 80 4.76 0.25 0.72 0.12 0.01 0.11 CORALIE – Udry et al. (in prep.) HD 168746 5610 30 4.50 0.15 1.02 0.08 −0.06 0.05 CORALIE 0.88 0.01 Pepe et al. (in prep.) HD 169830 6300 30 4.04 0.20 1.37 0.07 0.22 0.05 CORALIE 1.42 0.01 Naef et al. (2001a) HD 190228 5360 40 4.02 0.10 1.12 0.08 −0.24 0.06 WHT 0.84 0.01 Sivan et al. (2001) HD 202206bd 5765 40 4.75 0.20 0.99 0.09 0.37 0.07 CORALIE 1.02 0.02 Udry et al. (in prep.) HD 210277 5575 30 4.44 0.10 1.12 0.08 0.23 0.05 FEROS 0.94 0.01 Marcy et al. (1999) HD 217107 5660 50 4.43 0.10 1.04 0.07 0.40 0.06 CORALIE HD 217107 5655 40 4.42 0.05 1.11 0.08 0.38 0.05 FEROS HD 217107(avg) 5660 4.42 1.01 0.39 – 0.97 0.05 Fischer et al. (1999)

line list for iron and different sources of atmospheric pa- “pollution” scenario. The impact on the planetary forma- rameters (spectroscopic vs. colours) – see Paper I. This tion and evolutionary models is discussed. particular point may introduce systematic errors, and one might expect that the difference between the two samples was simply due to a bias related to the method used to 2. The data compute the metallicities. 2.1. The samples As mentioned above, the choice of a “comparison sam- With the goal of settling the question about the high ple” (as we shall call it from now on), free from giant metallicity content of stars with planets, we present here a planets and with the minimum bias possible, is of crucial spectroscopic study of a volume-limited sample of 43 stars importance. In order to be sure that we are not including included in the CORALIE (Udry et al. 2000) planet search any systematics when selecting the sample, we have taken programme, and for which the radial velocities seem to be all stars within the CORALIE sample (defined as volume- constant over a large time interval. The technique used, limited – Udry et al. 2000) having right ascensions between line lists and atmospheric models were those usually ap- 20h and 9h (observable in mid November from La Silla). plied by most authors working on the metallicities of stars Since such a sample corresponds to more than half of the with planets (e.g. Paper I; Gonzalez et al. 2001). We show ∼1600 stars searched for planets, it is virtually impossible that the currently known stars with giant planets are on to make a precise spectroscopic analysis for all those stars average more metal-rich than “field stars”, for which there in a reasonable amount time. We thus decided as a first is no radial-velocity signature of planets. Furthermore, the step to limit our observations to all stars within 17 pc of results are used to set strong constraints on the cause the Sun having (B−V ) < 1.1 (determined from Hipparcos of the observed “anomaly”, significantly excluding the data – ESA 1997), and within 20 pc with (B − V ) < 0.9. 1022 N. C. Santos et al.: The metal-rich nature of stars with planets

Table 3. Stars with giant planets or brown dwarf companions (bd) studied by other authors.

Star Teﬀ log gξt [Fe/H] Mass Reference Planet −1 (k) (cgs) (km s )[M ] for spectroscopy discovery paper

HD 9826 6140 4.12 1.35 0.12 1.28 0.02 Gonzalez & Laws (2000) Butler et al. (1997) HD 10697 5605 3.96 0.95 0.16 1.10 0.01 Gonzalez et al. (2001) Vogt et al. (2000) HD 12661 5714 4.45 0.99 0.35 1.01 0.02 Gonzalez et al. (2001) Fisher et al. (2000) HD 37124 5532 4.56 0.85 −0.41 – Gonzalez et al. (2001) Vogt et al. (2000) HD 46375 5250 4.44 0.80 0.21 – Gonzalez et al. (2001) Marcy et al. (2000) HD 75732A 5250 4.40 0.80 0.45 1.05 0.03 Gonzalez & Vanture (1998) Butler et al. (1997) HD 80606 5645 4.50 0.81 0.43 ∼1.1 Naef et al. (2001b) Naef et al. (2001b) HD 89744 6338 4.17 1.55 0.30 1.55 0.03 Gonzalez et al. (2001) Korzennik et al. (2000) HD 92788 5775 4.45 1.00 0.31 1.05 0.02 Gonzalez et al. (2001) Queloz et al. (2001) Fisher et al. (2000) HD 95128 5800 4.25 1.0 0.01 1.03 0.03 Gonzalez (1998) Butler & Marcy (1996) HD 114762bd 5950 4.45 1.0 −0.60 0.82 0.03 Gonzalez (1998) Latham et al. (1989) HD 117176 5500 3.90 1.0 −0.03 1.10 0.02 Gonzalez (1998) Marcy & Butler (1996) HD 120136 6420 4.18 1.25 0.32 1.34 0.02 Gonzalez & Laws (2000) Butler et al. (1997) HD 130322 5410 4.47 0.95 0.05 ∼0.89 Gonzalez et al. (2001) Udry et al. (2000) +0.07 HD 134987 5715 4.33 1.00 0.32 1.02−0.03 Gonzalez et al. (2001) Vogt et al. (2000) HD 143761 5750 4.10 1.2 −0.29 0.96 0.03 Gonzalez (1998) Noyes et al. (1997) HD 145675 5300 4.27 0.80 0.50 ∼1.05 Gonzalez et al. (1999) Udry et al (2001) HD 168443 5555 4.10 0.90 0.10 1.01 0.02 Gonzalez et al. (2001) Marcy et al. (1999) Udry et al. (2001) HD 177830 4818 3.32 0.97 0.36 1.03 0.11 Gonzalez et al. (2001) Vogt et al. (2000) HD 186427 5685 4.26 0.80 0.07 0.97 0.03 Laws & Gonzalez (2001) Cochran et al. (1997) HD 187123 5830 4.40 1.00 0.16 1.08 0.04 Gonzalez et al. (1999) Butler et al (1998) HD 192263 4964 4.49 0.95 −0.03 ∼0.80 Gonzalez et al. (2001) Santos et al. (2000b) Vogt et al. (2000) HD 209458 6063 4.38 1.02 0.04 1.12 0.02 Gonzalez et al. (2001) Mazeh et al. (2000) Henry et al. (2000) HD 217014 5795 4.41 1.05 0.21 1.07 0.01 Gonzalez et al. (2001) Mayor & Queloz (1995) HD 222582 5735 4.26 0.95 0.02 0.95 0.01 Gonzalez et al. (2001) Vogt et al. (2000)

This sample includes about 50 stars, most of which were 2.2. Observations and data reduction observed. Some of the stars in this sample have known planetary systems, namely HD 1237, HD 13445, HD 17051, For the “comparison sample”, spectra of S/N between HD 22049 and HD 217107 (HD 192263, would also have 150 and 350 were obtained using the CORALIE high- been included if we had not imposed any limit on (B−V )). resolution spectrograph (R ≡ ∆λ/λ ∼ 50 000), at the All these objects were excluded from our ﬁnal sample (43 Swiss 1.2-m Euler Swiss Telescope (La Silla, Chile) in objects) presented in Table 1. 2000 mid November. The spectra were reduced using IRAF2 tools, and the equivalent widths were measured by ﬁtting a Gaussian function to each of the lines using In what concerns the stars with planets, all known ob- the “k” key within splot. jects having precise spectroscopic iron abundance determi- Some planet host stars were also observed with nations using a similar technique were used. This excludes CORALIE (adding a few objects to the study presented Gl 876 (Delfosse et al. 1998; Marcy et al. 1998), HD 195019 in Paper I), some of them with no previous spectroscopic (Fischer et al. 1999) and the recently announced plan- abundance determinations (see Table 2). Observations of ets around HD 160691, HD 27442 and HD 179949 (Tinney some stars with planets were also complemented with et al. 2001; Butler et al. 2001). Also, from the set of spectra of S/N ∼ 300 taken with the FEROS spec- 9 new planet host stars recently announced by the Geneva trograph (R ∼ 48 000), at the ESO 1.52 m Telescope group (Mayor et al. 2001b) only two had available data (La Silla), during the nights of 2000 November 8 and 9. (HD 28185 and HD 80606). The only exception was made The UES spectrograph (R ∼ 55 000) at the 4 m WHT respecting BD−10 3166, which was included in the planet surveys for its high [Fe/H] (Gonzalez et al. 1999). All the other stars were part of search programmes for planets in 2 IRAF is distributed by National Optical Astronomy volume-limited samples (Udry et al. 2000; Marcy et al. Observatories, operated by the Association of Universities for 2000), and there is no particular reason to take out any of Research in Astronomy, Inc., under contract with the National the targets (possible sources of bias are discussed below). Science Foundation, USA. N. C. Santos et al.: The metal-rich nature of stars with planets 1023

Fig. 1. Left: distribution of stars with planets (shaded histogram) compared with the distribution of ﬁeld dwarfs presented in this paper (empty histogram). The dashed histogram represents the planet host sample if we consider only stars forming part of the CORALIE survey (see text for details). The vertical lines represent stars with brown dwarf candidate companions having minimum masses between 10 and 20 MJup. Right: The cumulative functions of both samples. A Kolmogorov–Smirnov test shows the probability of the stars’ being part of the same sample is around 10−7.

(La Palma, Canary Islands) was also used to obtain a from spectroscopy3. Only for HD 222237 we do not present spectrum for the planet host star HD 190228. a mass, given the high error in its determination. In Table 2 we present our results of the spectroscopic analysis for stars with low mass companions for which we 2.3. Spectroscopic analysis have obtained spectra. Also included are the objects already presented in Paper I and Santos et al. (2001), since The technique used has already been described extensively the values were revised; the differences are always per- in Paper I. Abundances and atmospheric parameters were fectly within the uncertainties (usually lower than 0.02 dex determined using a standard local thermodynamic equi- for [Fe/H]). Errors in the parameters were computed as librium (LTE) analysis with a revised version of the line described in Paper I. abundance code MOOG (Sneden 1973), and a grid of For a few stars in our sample we had both CORALIE Kurucz (1993) ATLAS9 atmospheres. As in Paper I, we and FEROS spectra. No important systematics are evi- used a set of iron lines taken from the list of Gonzalez dent in the results from the two spectrographs (due to & Laws (2000). Here, we added one more Fe ii line (the errors in flatfields or background subtraction), and we sim- 5991 A˚ presented in Gonzalez et al. 2001) and we excluded ply used a mean value of the results. the previously used Fe ii line at 6432 A˚ line, because it was We compared our [Fe/H] determinations for the eight giving systematically low values for most stars. stars presented in this paper in common with the studies of Gonzalez et al. (1999), Gonzalez & Laws (2000), The atmospheric parameters were obtained from the and Gonzalez et al. (2001) to look for any possible sys- Fe i and Fe ii lines by iterating until the correlation coef- tematics. The stars and [Fe/H] obtained by these authors ficients between log (Fe i)andχl, and between log (Fe i) are HD 1237 (0.16), HD 16141 (0.15), HD 17051 (0.19), and log (Wλ/λ) were zero, and the mean abundance given HD 38529 (0.37), HD 52265 (0.27), HD 75289 (0.28), by Fe i and Fe ii lines were the same. This procedure gives HD 210277 (0.24) and HD 217107 (0.36). The comparison very good results since the set of Fe i lines has a very wide shows that the formal mean difference in [Fe/H] obtained range of excitation potentials. We used log (Fe) = 7.47. by these authors and from the present work is virtually 0.00 dex; the rms of the differences is ∼0.01 dex. The at- The complete results obtained for the comparison sam- mospheric parameters for these stars are also usually the ple are presented in Table 1. The errors in T ,logg, ξ eff t same within the errors. We thus conclude that we can in- and [Fe/H] for a typical measure are of the order of 50 K, clude their determinations for stars with planets in our 0.15 dex, 0.10 dex, and 0.06 dex respectively. The number analysis (as expected since our studies were similar in all of lines used for each star, and the dispersions for Fe i aspects to theirs) without introducing any systematics, and Fe ii lines are also tabulated. The masses were deter- thus increasing the number of available planetary host mined from theoretical isochrones of Schaller et al. (1992), 3 Schaerer et al. (1992) and Schaerer et al. (1993), using MV The errors in mass represent simply formal errors, com- computed from Hipparcos parallaxes and Teff obtained puted using the uncertainties in Teff and MV . 1024 N. C. Santos et al.: The metal-rich nature of stars with planets

volume-limited samples of stars. The only exception is BD−10 3166 (Butler et al. 2000), chosen for its high metallicity. This star was not included in our analysis. It is worthy of note that the five planet-host stars that were included in our volume-limited sample (HD 1237, HD 13445, HD 17051, HD 22049 and HD 217107) have a mean [Fe/H] of +0.10. No important systematics are expected concerning the magnitudes of the objects. On the one hand, for a given colour higher [Fe/H] stars are more luminous. Also, higher metallicity implies more and deeper lines, and thus a more precise determination of the velocity. But a star with more metals is also, for a given mass, cooler and fainter. For ex- ample, doubling the metallicity of the Sun (i.e. increasing [Fe/H] to ∼0.30) would make its temperature decrease by more than 150 K, and its luminosity by a factor of ∼1.2 (Schaller et al. 1992; Schaerer et al. 1993). As we shall see below, for the mass intervals for which we have a good representation of both samples, stars with planets are always significantly more metal-rich. Furthermore, at least in the CORALIE survey, exposure times are computed in Fig. 2. The shape of the distribution of metallicities for stars order to have a photon-noise error at least as low as the with planets increases with [Fe/H]. This represents also an in- instrumental errors. crease with Z, the mass fraction of heavy elements. Given the uniformity of the study, we may conclude that the plot in Fig. 1 represents a proof that the stars now observed with giant planets are, on average, more metal- candidates used in the current study. The list of stars with rich than field stars. For the record, it is also interesting to planets whose determinations were made by other authors verify that the mean value of [Fe/H] obtained by Favata is presented in Table 3. et al. (1997) for their volume-limited sample of G-dwarfs is −0.12, which means that there are almost no systematics 3. The metallicity distributions between the method used by these authors and that used for the current study. In Fig. 1 we present the [Fe/H] distributions of both sam- Also remarkable in Fig. 1 is the interesting position of ples described above. For the stars with planets we in- the three low-mass (10 MJup

Fig. 3. Panel a): Metallicity distribution of stars with planets (dashed histogram) compared with the distribution of a volume limited sample of field dwarfs (empty histogram) – see text for more details; panel b): correcting the distribution of stars with planets from the same distribution for stars in the volume-limited sample results in a even more steep rise of the planet host star distribution as a function of [Fe/H]. Note that the last bin in the planet distribution has no counterpart in the field star distribution; panel c):sameasa) but for a field distribution computed using a calibration of the CORALIE cross-correlation dip to obtain the metallicity for ∼1000 stars; panel d):sameasb) when correcting from the field distribution presented in c). this distribution is well described by a quadratic rising disappears, will increase quadratically with Z (considering with [Fe/H] (up to [Fe/H] ∼ +0.35 – dashed curve in the that there is a linear relation between these two variables). diagram). In other words, the distribution is rising with Z But this model is probably too simple. Other, more (the mass fraction of heavy elements) up to a value of 0.04, complicated mechanisms intervene, however, such as the falling then abruptly. This sharp cut-off in the distribution fact that an increase in [Fe/H] will also probably increase of stars with planets is probably due to the fact that we the opacity of the disc, changing its physical conditions may be looking at a limit on the metallicity of the stars (e.g. Schmidt et al. 1997), such as the accretion rate or in the solar neighbourhood. its vertical structure. The current result may be used in exactly this way in constraining these mechanisms. Clearly, the rise is expected if we consider that the On the other hand, the distribution of [Fe/H] in a probability of forming planets is proportional to the metal- volume-limited sample of stars (with and without planet licity content. In one sense, an increase in the number of hosts) apparently decreases with increasing [Fe/H] for dust particles (N) in the young disc will increase the prob- [Fe/H] > 0.0 (e.g. Favata et al. 1997). Taking this ef- ability of a shock by N 2. The rate of formation of planetes- fect into account would result in an even steeper rise imals, and thus the probability of creating a planet core of the relative frequency of stars with planets with the with sufficient mass to accrete gas from the disc before it metallicity. 1026 N. C. Santos et al.: The metal-rich nature of stars with planets

Fig. 4. Distribution for the ﬁeld star sample (empty histogram) if we add 15 earth masses of iron (left) or a quantity of iron proportional to Z (right) – in the latter case, 0.003×Z. The distributions peak at about the same value as the distribution of stars with planets (dashed histogram), but have completely distinct forms and extents in [Fe/H]. See the text for more details.

This is exactly what we can see from Fig. 3. In pan- of [Fe/H] > 0.0. Although this is a preliminary result, els (a) and (c) we present a comparison between the the plots leave no doubts about the strong significance of [Fe/H] distribution of stars with planets (shaded bars) and the increase in the frequency of known planetary systems the same distribution for two volume-limited samples of found with the metallicity content of their host stars. dwarfs (empty bars). The volume-limited distribution in We note that for the low metallicity tail of the dis- panel (a) was taken from this paper (objects presented tribution, this result is based on relatively poor statistics. in Table 1 plus the five stars with planets making part The small relative number of stars with planets found with of our volume limited sample – see Sect. 2.1 for details). low [Fe/H] values is, however, perfectely consistent with For panel (c), the distribution represents the metallicities recent studies that failed to find any planetary host com- of about 1000 stars in the CORALIE planet-search sam- panions amid the stars in the metal-poor globular cluster ple as determined from a calibration of the CORALIE 47 Tuc (Gilliland et al. 2000). cross-correlation function surface (Santos et al., in prepa- ration)5. Since these two field star distributions represent the actual distribution of metallicities in the solar neigh- 4.2. A simple “pollution” model borhood (within statistical errors), we can use them to We tried to determine whether it would be possible to ob- estimate the “real” percentage of stars with planets for tain a metallicity distribution for the stars without planets each metallicity bin that we would find if the metallic- similar to that of stars with extrasolar planets by simply ity distribution of field stars was “flat” – panels (b) and adding material to their convective envelopes. To make (d). As we can see from these plots, there is a steep rise this simple model, we computed the convective zone mass of the percentage of stars with planets with [Fe/H] (we for each star using Eqs. (5) and (6) of Murray et al. (2001) will concentrate on panel (d) given that it presents bet- – metallicity-dependent relations for M<1.2 M ;only ter statistics than panel (b)). Actually, the plot shows stars in this mass range were used, since the comparison that more than 75% of the stars with planets would have sample is not complete for higher masses. We are suppos- [Fe/H] > +0.2 dex. But even more striking is the fact that ing that the “pollution”, if significant, has taken place af- ∼ 85% of the surface of the distribution falls in the region ter the star reaches the ZAMS. As discussed in Paper I, if we consider the pollution to occur in a pre-main-sequence 5 The cross-correlation dip can be seen as an average spectral phase, higher quantities of material would be needed, since line; as shown by Mayor (1980) and Pont (1997) its surface is the convection envelope masses are bigger (D’Antona & well correlated with [Fe/H] and (B − V ). A calibration of its Mazzitelli 1994). In fact, recent results show that gas-disc surface with these two variables can in fact be used to obtain lifetimes can be higher than previously thought (up to in a very simple way precise metallicity estimates, at least in a 20 Myr – Thi et al. 2001), and thus we can probably ex- statistical sense. In this particular case, the calibration used to pect significant pollution when the star is already on the construct this “comparison” sample uses as “calibrators” the stars presented in Tables 1 and 2, and thus the results are in the main sequence. same [Fe/H] scale as our spectroscopic values. The precision of Two models were then constructed: the first was built the calibration is remarkable (rms ∼ 0.01 dex). The resulting on the supposition that the quantity of material falling distribution is quite symmetric, and peeks at ∼−0.05 dex. into the star is the same for all the objects. A second, and N. C. Santos et al.: The metal-rich nature of stars with planets 1027

4.3. The mass dependence It has been suggested (Laughlin & Adams 1997) that if the pollution scenario is correct one would be able to see a trend of [Fe/H] with stellar mass, since the high-mass stars have tiny convective envelopes, and thus a higher probability of having a noticeable increase in [Fe/H] (here we have chosen stellar mass and not the mass of the convective shell, since if the pollution scenario is correct, the original mass for the convection zone for the stars with planets would not be correctly calculated). In Fig. 5 we present such a plot. There are three features in the plot that deserve a comment. First, the region for M ≥ 1.2 M has a very low number of stars without planets. This is because in a volume-limited sample of stars, there are very few dwarfs in this mass regime (the IMF decreases with mass). However, these stars are bright, and thus easier targets for planet searches, which explains the high number of planetary candidates in this region. On the other hand, for Fig. 5. A plot of [Fe/H] as a function of mass for planet host masses M ≤ 0.8 M , we see the opposite effect: stars stars (filled circles) and “comparison” stars (open squares). are fainter, and thus difficult targets for planet searches, The line represents the upper limit for the field star [Fe/H]. but are more numerous in a volume-limited sample. There See text for more details. are probably, however, two more biases in the diagram. In fact, the comparison sample was cut in (B − V ). For a given temperature, a higher-metallicity star also has higher mass. Thus, stars in the upper part of the dia- possibly more realistic model, considered the pollution to gram (more metal-rich) are moved to the righthand side be proportional to Z. The quantity of “added material” relative to stars in the lower part. For a fixed Teff ,0.2dex was chosen so that the distributions (stars with and within metallicity imply ∼0.05 M . On the other hand, if we out planets) peak at about the same value of [Fe/H]. take a given volume-limited sample of dwarfs, lower-mass The results are shown in Fig. 4 (open bars), compared dwarfs tend to be older than high-mass dwarfs; in propor- with the results obtained for stars with planets within the tion there are thus fewer metal-rich dwarfs in the “high”- same mass regime. As we can see, the shape of the planet mass region of the plot. These facts might explain the sample, particularly the sharp cutoff at high metallicity, is lack of stars with planetary companions presenting M ≥ not correctly obtained for any of the models. The result- 1.2 M and low [Fe/H]. ing samples are more symmetric, but most of all, in both In any case, the most striking thing in the diagram is models there are always a few points that end up with the fact that all comparison stars have [Fe/H] ≤ 0.17 dex, extremely high metallicities (>0.8 dex). This is particu- while about 55% of stars with planets have metallicity larly unpleasant, since the inclusion of higher-mass stars above that value. Furthermore, contrary to former studies (such as some of the observed exoplanet hosts) would even (Laughlin 2000) this diagram does not suggest that the boost some bins to higher values, given that the mass metallicity of the upper envelope of stars with planets is of the convection zone is supposed to be lower. As dis- increasing with stellar mass. This point renders unlikely cussed by Murray et al. (2001), however, Li and Be ob- the possibility that the high metallicity of these stars is servations suggest that the total “mixing” mass increases due to pollution. As discussed above, however, for stel- after 1.2 M , eventually reducing this problem. Note also lar masses higher than ∼1.2 M , the total “mixing” mass that the strong cut-off in [Fe/H] for the planet host star may increase. This may explain to some extent the lower sample remains, even though this time we have not in- value of the upper envelope of [Fe/H] for masses above cluded stars with M>1.2 M . This shape is not at all this limit, but it is not certain that it is enough to explain reproduced by our computations. the difference. An extension of our comparison sample for We caution, however, that we are using very simple stars with M>1.2 M would be very useful. models to make this calculation. The inclusion of more As mentioned, our results and interpretations are op- parameters, such as a time- or disc-mass-dependent ac- posed to those presented by Laughlin (2000), who used cretion (which might be related to the stellar mass), or photometric indices to compute the metallicities for a sam- a mass-dependent size of the central disc hole (e.g. Udry ple of stars with planets, and for a large volume-limited et al. 2001), might modify the results. In any case, we have sample of dwarfs used as “comparison”. We believe that shown that a simple “pollution” model cannot explain the there are two important reasons for the difference. First, observations. the trend he finds in [Fe/H] vs. stellar mass is mainly 1028 N. C. Santos et al.: The metal-rich nature of stars with planets

Table 4. Planetary and orbital parameters used in Figs. 6 and 0.21 dex, for a mean stellar mass of 0.95 M . To pollute 7. Stars with multiple companions are not included. The values the convective envelope of a 0.95 solar mass star contain- were compiled from the literature. ing about 0.04 M in the convection zone (D’Antona & Mazzitelli 1994) in order to obtain the observed differ- HD Star M sin i Period ae ence, one would need about 10 earth masses of pure iron. number (MJup)(days)(AU) Considering the composition of C1 chondrites, this would mean ∼5 times more in silicate material, a value that 1237 GJ 3021 3.51 133.7 0.494 0.51 seems excessively high. The addition of giant planets, also 6434 HD 6434 0.48 22.1 0.15 0.295 very rich in gas, would need even more material. For exam- 10697 109 Psc 6.60 1072 2.12 0.12 12661 HD 12661 2.93 264.0 0.789 0.33 ple, the addition of Jupiter to the present-day Sun would ∼ 13445 Gl 86 3.77 15.8 0.11 0.04 only be able to increase its metallicity by 0.05 dex, while 16141 HD 16141 0.23 75.8 0.35 0.28 adding two jupiters would increase 0.08 dex (e.g. Gonzalez 17051 ι Hor 2.36 311.3 0.93 0.22 1998). We are supposing here that Jupiter has a rocky 19994 HD 19994 1.88 454.2 1.23 0.2 core, which is not necessarily true (Guillot 1999). 22049 Eri 0.84 2518 3.4 0.6 Murray et al. (2001) have reported evidence that all 28185 HD 28185 5.6 385 1.0 0.06 stars in the solar neighbourhood have accreted about 37124 HD 37124 1.04 154.8 0.55 0.31 0.4 Earth masses of iron, suggesting that “terrestrial- 38529 HD 38529 0.81 14.3 0.129 0.27 type material is common around solar type stars”. This 46375 HD 46375 0.26 3.02 0.041 0.02 seems to be particularly evident in the objects with M> 52265 HD 52265 1.08 119.2 0.5 0.35 1.3–1.4 M , which is in accordance with expectations 75289 HD 75289 0.44 3.48 0.04 0.065 75732A 55 Cnc 0.82 14.66 0.12 0.03 (Laughlin et al. 1997). If confirmed, and as discussed by 80606 HD 80606 3.51 111.8 0.44 0.93 these authors, this result further stresses the difficulty of 89744 HD 89744 7.54 265.0 0.91 0.7 obtaining the [Fe/H] differences we observe considering a 92788 HD 92788 3.77 340.8 0.97 0.36 pollution model: 0.4 Earth masses of iron are definitely 95128 47 UMa 2.57 1084 2.1 0.13 not able to explain the observed differences between stars 108147 HD 108147 0.35 11.05 0.10 0.57 with planets and stars “without” planets. 117176 70 Vir 7.75 116.7 0.48 0.4 120136 τ Boo 4.29 3.31 0.047 0.02 121504 HD 121504 0.89 64.62 0.32 0.13 4.4. Using evolved stars as a test 130322 HD 130322 1.04 10.72 0.088 0.044 One other interesting note can be added by looking at the 134987 23 Lib 1.57 259.6 0.81 0.24 143761 ρ CrB 1.15 39.6 0.23 0.07 “star-with-planet” sample. If the pollution scenario were 145675 14 Her 5.65 1650 2.84 0.37 correct, we would expect evolved stars (sub-giants) hav- 168746 HD 168746 0.25 6.41 0.066 0.00 ing planetary companions to have lower metallicities than 169830 HD 169830 3.04 229.9 0.82 0.35 their dwarf counterparts, since the convection envelope in- 177830 HD 177830 1.26 391.6 1.1 0.41 creases in mass as the star evolves off the main sequence 186427 16 Cyg B 1.75 804.4 1.61 0.67 (Sackmann et al. 1993), diluting its metallicity content. 187123 HD 187123 0.60 3.10 0.042 0.01 There are eight stars in Tables 2 and 3 having spec- 190228 HD 190228 5.03 1161 2.3 0.5 troscopic log g values below ∼4.10 dex (probably already 192263 HD 192263 0.75 24.13 0.15 0.00 slightly evolving away from the main sequence): HD 10697 209458 HD 209458 0.68 3.52 0.047 0.00 (3.96), HD 38529 (4.01), HD 117176 (3.90), HD 143761 210277 HD 210277 1.31 435.6 1.09 0.34 (4.10), HD 168443 (4.10), HD 169830 (4.04), HD 177830 217014 51 Peg 0.47 4.23 0.05 0.00 (3.32) and HD 190228 (4.02). If we compute the mean 217107 HD 217107 1.34 7.11 0.071 0.14 222582 HD 222582 5.44 575.9 1.35 0.71 [Fe/H] we obtain a value of +0.08 0.25 (with values going from −0.29 to +0.39), only slightly lower than the value of +0.15 0.22 for the sample comprising only stars with log g>4.10 (values going from −0.55 to +0.50). As due to the few “low”-[Fe/H] planet host stars having mentioned also by Gonzalez et al. (2001), this is even more M<1 M . On the other hand, in his plot there are a striking when we notice that the higher-metallicity star in few very high metallicity “comparison” stars in the region this “low log g” sample (HD 177830) is that having the of M<1.1 M . If we believe in the results presented in lower surface gravity. this paper, the probability that these stars harbour planets is very high. His sample might in fact be biased by the 5. Metallicity and orbital parameters presence of planet host stars in his “comparison” sample. If we concentrate on the mass interval for which we Given that we already have more than 40 extrasolar plan- have both stars with planets and comparison stars (region ets with high-precision metallicity determinations, we can with 1.2 M >M>0.7 M ), we obtain a mean [Fe/H] = start to think about looking for possible trends in [Fe/H] +0.13 for the star with planet sample, and −0.08 for the with planetary mass, semi-major axis or period, and comparison star sample. This makes a difference of about eccentricity. N. C. Santos et al.: The metal-rich nature of stars with planets 1029

Fig. 6. Left: distribution of [Fe/H] for stars with planets orbiting with semi-major axes lower than 0.1 AU (shaded histogram), and with semi-major axes greater than 0.1 AU (open histogram). The vertical lines denote stars with more than one planetary- mass companion. Right: cumulative fraction of the [Fe/H] distribution for the long (thin line) and short (thick line) period systems. Althought some trend is suggested regarding a possible higher metallicity for the stars with short period planets, a Kolmogorov–Smirnov test gives a probability of ∼0.7, which is not signiﬁcant.

Gonzalez (1998) and Queloz et al. (2000) have sug- precise abundances) are needed to permit conclusions. In gested that stars with short-period planets (i.e. small the same way, no conclusions are possible regarding the semi-major axes) may be particularly metal-rich, even position of the planetary systems in the plot (Fig. 6). amongst the planetary hosts. The number of planets that were known by that time was, however, not enough to 6. Concluding remarks arrive at a definitive conclusion. Here we revisit that dis- cussion with much better statistics. We have presented for the first time a uniform and un- In Fig. 6 we present a plot of the distribution of planets biased comparison between the metallicity distributions for two orbital “radius” regimes: stars with planets orbit- of a sample of stars with planets and a sample of stars ing with semi-major axes lower than 0.1 AU (shaded his- “without” planets. The main results can be summarized togram), and with semi-major axes greater than 0.1 A.U as follows: (open histogram). The values were taken from Table 4. Stars with multiple companions are denoted by the verti- – The currently known stars with planets are substan- cal lines and were not included in the histograms. These tially metal-rich when compared with non-planetary ∼ include, among the stars for which we have [Fe/H] values host dwarfs. The mean difference in [Fe/H] is 0.25 dex available, the multiple systems around υ And (Butler et al. and is clearly significant. 1999), HD 83443 (Mayor et al. 2001a), HD 168443 (Udry – The shape of the metallicity distribution of stars with et al. 2001), and HD 82943 (Mayor et al. 2001b). planets has a very clear rise with [Fe/H], i.e. with Z, the mass fraction of heavy elements. This result, al- Some trend can be seen in the plot (right panel) sug- though still based upon “only” 44 objects, may help gesting that stars with short period planets might be put constraints on the models and conditions leading more metal-rich then their long-period counterparts. A to giant-planet formation. Kolmogorov–Smirnov test, however, gives a probability – The clear cut-off seen for the star-with-planet distribu- that both distributions make up part of the same sam- tion at [Fe/H] ≥ +0.5 may be interpreted as a “limit” ∼ ple of 75%. We have tried to make the plot using differ- for the metallicity of solar-type dwarf stars in the so- ent limits of the semi-major axes; no further conclusions lar neighbourhood; it is very difficult to explain it in can be drawn. It is interesting to note that one of the terms of a convective envelope “pollution” scenario. longest-period planets discovered up to now (around the – The metallicity of stars with planets cannot be ex- star 14 Her) has a value of [Fe/H] = +0.50. plained by a simple “pollution” model. The differ- Similar conclusions are drawn respecting the eccen- ences in [Fe/H] between stars with planets and field tricity or the companion mass (Fig. 7). The current re- stars does not seem to be explained by such mecha- sults do not suggest any clear difference in [Fe/H] be- nisms, but needs an original (cloud) metallicity “ex- tween high- and low-mass/eccentricity planets. If there cess”. Otherwise huge quantities of iron-rich material is some trend, then more data (and probably even more would have to be engulfed, which is not easy to explain 1030 N. C. Santos et al.: The metal-rich nature of stars with planets

Fig. 7. Left: distribution of [Fe/H] for stars with planets having masses lower than 1 MJup (shaded histogram), and greater than 1 MJup (open histogram). Right: Similar to the left panel, but for eccentricity (limit at 0.1). In both cases, a Kolmogorov-Smirnov test results in a non-signiﬁcative probability of being two diﬀerent populations.

in face of the current observations. Our results do not not change in any way the conclusions of this paper, since rule out, however, planet accretion as an ongoing pro- their models predict “pollution” of the order of 0.4 Earth cess in the early stages of planetary formation (during masses of iron, only noticeable in stars with M>1.5 M . which the convective envelopes are more massive). One remarkable point that becames evident from the – An analysis of the planetary orbital parameters does current work is the fact that our Sun occupies a “modest” not reveal any clear trends with [Fe/H]. This point position in the low [Fe/H] tail of the metallicity distribu- might give further support to the “primordial” ori- tion of stars with planets. A look at the orbital parame- gin of the high metallicity of stars with planets (e.g. ters in Table 4, however, tells us that, to date, no “real” Gonzalez 1998; Del Popolo et al. 2001). On the other Solar System analogs were found. This lead us to specu- hand, we cannot exclude the possibility that some late about possible different formation histories for these trend may appear in the future, e.g. when some real systems. Note, however, that we cannot draw any conclu- solar system analogues are found. sions until other Solar System analogs are found. The future addition of more stars in both samples stud- The current result concerning the high metallicity of stars ied here, and the detailed analysis of other chemical ele- with planets may represent the simple fact that the higher ments (in particular the light elements Li, Be and B) are the metallicity of the star, the higher will be the proba- of particular interest for the future, both to consolidate bility that planet formation will occur. current results, and to understand better the multitude of Recently, Israelian et al. (2001) have discovered that different planetary systems known today. The direct com- the planet host star HD 82943 has 6Li in its atmosphere. parison of other elemental abundances (e.g. α-elements, This discovery is interpreted as evidence of the infall of C, or O) between stars with planets and “single” field a planet into the star, most probably by planet disrup- dwarfs will, on the other hand, not be an easy task given tion (Rasio & Ford 1996). As discussed by the authors, the apparent lack of stars with high [Fe/H] having no giant however, the high [Fe/H] value of this star is not neces- planet companions. Meanwhile, we expect that theoretical sarily related to the infall of a planet, as this would not models of disc evolution and (giant) planetary formation be able to enhance the [Fe/H] value by more than a few will be constructed to explain in a detailed way the physics hundredths of a dex. This is exactly what was found for beyond the current results. the pair 16 Cyg A and B (Laws & Gonzalez 2001). In fact, the present results do not exclude the possibility that pollution may play a role (eventually important Acknowledgements. We wish to thank the Swiss National in some cases, such as for the most massive dwarfs), but Science Foundation (FNSRS) for the continuous support for rather that it is not the key process leading to the observed this project. We would also like to thank Nami Mowlavi for high metallicity of the planet host stars. Recent work by help computing masses for the stars, and to Terry Mahoney Murray et al. (2001) supports the idea that pollution, on for the useful english comments. Support from Funda¸cão para a small scale, is very common among the stars of the so- aCiência e Tecnologia, Portugal, to N.C.S. in the form of a lar neighbourhood. If confirmed, however, this result does scholarship is gratefully acknowledged. N. C. Santos et al.: The metal-rich nature of stars with planets 1031

References Marcy, G. W., Butler, R. P., Vogt, S., et al. 1999, ApJ, 520, 239 Boss, A. P. 2000, ApJ, 536, L101 Marcy, G. W., Butler, R. P., & Vogt, S. 1998, ApJL, 505, 147 Butler, R. P., et al. 2001, ApJ, submitted Marcy, G. W., & Butler, R. P. 1999, ApJL, 464, 147 Butler, R. P., Vogt, S. S., Marcy, G. W., et al. 2000, ApJ, 545, Mayor, M., Naef, D., Pepe, F., et al. 2001a, in Planetary 504 Systems in the Universe: Observations, Formation and Butler, R. P., Marcy, G., Fisher, D., et al. 1999, ApJ, 526, 916 Evolution, ed. A. Penny, P. Artymowicz, A.-M. Lagrange, Butler, R. P., Marcy, G., Vogt, S., & Apps, K. 1998, PASP, & S. Russel, ASP Conf. Ser., in press 110, 1389 Mayor et al. 2001b, ESO Press-Release 07/01 (www.eso.org/ Butler, R. P., Marcy, G. W., Williams, E., et al. 1997, ApJ, outreach/press-rel/pr-2001/pr-07-01.html) 474, L115 Mayor, M., & Queloz, D. 1995, Nature, 378, 355 Butler, R. P., & Marcy, G. W. 1996, ApJ, 464, L153 Mayor, M. 1980, A&A, 87, L1 Cochran, W., Hatzes, A., Butler, R. P., et al. 1997, ApJ, 483, Murray, M., Chaboyer, B., Arras, P., et al. 2001, submitted to 457 ApJ D’Antona, F., & Mazzitelli, I. 1994, ApJS, 90, 467 Naef, D., Mayor, M., Pepe, F., et al. 2001a, A&A, submitted Delfosse, X., Forveille, T., Mayor, M., et al. 1998, A&A, 338, Naef, D., Latham, D., Mayor, M., et al. 2001b, A&A, submitted L67 Noyes, R., Jha, S., Korzennik, S., et al. 1997, ApJL, 483, 111 Del Popolo, A., Gambera, M., & Nihal Ercan, E. 2001, Pont, F. 1997, Ph.D. Thesis, University of Geneva MNRAS, in press Queloz, D., Mayor, M., Naef, D., et al. 2001, in Planetary ESA, 1997, The HIPPARCOS and TYCHO catalogue, Systems in the Universe: Observations, Formation and ESA-SP 1200 Evolution, ed. A. Penny, P. Artymowicz, A.-M. Lagrange, Favata, F., Micela, G., & Sciortino, S. 1997, A&A, 323, 809 & S. Russel, ASP Conf. Ser., in press Fischer, D., Marcy, G. W., Butler, R. P., et al. 2000, ApJ, Queloz, D., Mayor, M., Weber, L., et al. 2000, A&A, 354, 99 submitted Rasio, F. A., & Ford, E. B. 1996, Science, 274, 954 Fischer, D., Marcy, G. W., Butler, R. P., et al. 1999, PASP, Sackmann, I.-J., Boothroyd, A. I., & Kraemer, K. E. 1993, 111, 50 ApJ, 418, 457 Gilliland, R. L., Brown, T. M., Guhathakurta, P., et al. 2000, Santos, N. C., Israelian, G., & Mayor, M. 2001, in Planetary ApJ, 545, L47 Systems in the Universe: Observations, Formation and Goldreich, P., & Tremaine, S. 1980, ApJ, 241, 425 Evolution, ed. A. Penny, P. Artymowicz, A.-M. Lagrange, Gonzalez, G., Laws, C., Tyagi, S., & Reddy, B. E. 2001, AJ, & S. Russel, ASP Conf. Ser., in press 121, 432 Santos, N. C., Israelian, G., & Mayor, M. 2000a, A&A, 363, Gonzalez, G., & Laws, C. 2000, AJ, 119, 390 228 (Paper I) Gonzalez, G., Wallerstein, G., & Saar, S. 1999, ApJL, 511, 111 Santos, N. C., Mayor, M., Naef, D., et al. 2000b, A&A, 356, Gonzalez, G. 1998, A&A, 334, 221 599 Gonzalez, G., & Vanture, A. D. 1998, A&A, 339, L29 Schaerer, D., Charbonnel, C., Meynet, G., et al. 1993, A&AS, Gonzalez, G. 1997, MNRAS, 285, 403 102, 339 Guillot, T. 1999, Science, 286, 72 Schaerer, D., Meynet, G., Maeder, A., et al. 1992, A&AS, 98, Hatzes, A. P., Cochran, W. D., McArthur, B., et al. 2000, 523 ApJL, 544, 145 Schaller, G., Schaerer, D., Meynet, G., & Maeder, A. 1992, Henry, G. W., Marcy, G. W., Butler, R. P., & Vogt, S. 2000, A&AS, 96, 269 ApJ, 529, L41 Schmidt, W., Henning, T., & Mucha, R. 1997, A&A, 325, 569 Israelian, G., Santos, N. C., Mayor, M., & Rebolo, R. 2001, Sivan, J.-P., Mayor, M., Naef, D., et al. 2001, in Planetary Nature, 411, 163 Systems in the Universe: Observations, Formation and Korzennik, S. G., Brown, T. M., Fisher, D. A., et al. 2000, Evolution, ed. A. Penny, P. Artymowicz, A.-M. Lagrange, ApJ, 533, L147 & S. Russel, ASP Conf. Ser., in press K¨urster, M., Endl, M., Els, S., et al. 2000, A&A, 53, L33 Smith, V. V., Cunha, C., & Lazzaro, D. 2001, AJ, in press Kurucz, R. L. 1993, CD-ROMs, ATLAS9 Stellar Atmospheres Sneden, C. 1973, Ph.D. Thesis, University of Texas Programs and 2 km s−1 Grid (Cambridge: Smithsonian Thi, W. F., Blake, G. A., Van Dishoeck, E. F., et al. 2001, Astrophys. Obs.) Nature, 409, 60 Latham, D. W., Stephanik, R. P., Mazeh, T., & Mayor, M. Tinney, C., Butler, R. P., Marcy, G., et al. 2001, ApJL, in press 1989, Nature, 339, 38 Udry, S., Mayor, M., & Queloz, D. 2001, in Planetary Systems Laws, C., & Gonzalez, G. 2001, ApJ, in press in the Universe: Observations, Formation and Evolution, Laughlin, G. 2000, ApJ, 545, 1064 ed.A.Penny,P.Artymowicz,A.-M.Lagrange,&S.Russel, Laughlin, G., & Adans, F. C. 1997, ApJ, 491, L51 ASP Conf. Ser., in press Lin, D., Bodenheimer, P., & Richardson, D. C. 1996, Nature, Udry, S., Mayor, M., Naef, D., et al. 2000, A&A, 356, 590 380, 606 Vogt, S. S., Marcy, G. W., Butler, R. P., & Apps, K. 2000, Lissauer, J. 1995, Icarus, 114, 217 ApJ, 536, 902 Mazeh, T., Naef, D., & Torres, G. 2000, ApJ, 532, L55 Marcy, G. W., Butler, R. P., & Vogt, S. 2000, ApJL, 536, 43